Why Ethnic Groups Rebel
Total Page:16
File Type:pdf, Size:1020Kb
Why Ethnic Groups Rebel: Intra-ethnic Division, Dynamic Grievances, State Repression and Escalation Keiichi KUBO Thesis submitted for the degree of PhD in Political Science London School of Economics and Political Science Department of Government London, 2007 1 \ UMI Number: U501706 All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. Dissertation Publishing UMI U501706 Published by ProQuest LLC 2014. Copyright in the Dissertation held by the Author. Microform Edition © ProQuest LLC. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code. ProQuest LLC 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106-1346 {■ 6? 6 4 - &A, \im *4- :btbH gqutnOyindl3 vri W t^anK79hO oimftn/fl nobr/id aindio-B'rtnl iio h d im 3 bne noi0«3iqs^l OflUi' ifbibX to osigoh 3ffj lot bstfjimdu* zi^riT ooivjio?. LeomloH ni Gricl oonob8 IsdrliloH brm .goimonooH r!'o loorb2 riobnoJ JnornrnovoO 'to tnomttBq >(I U)()i nobno.l Library jjrii& rrnr.'i *: '■” • Declaration I hereby certify that the work presented in the thesis is solely my own work other than where I have clearly indicated that it is the work of others. Tokyo, 30 November, 2007 KeiichiKUBO 2 Abstract Why do ethnic groups rebel against the state? While there have been various efforts to answer this question, existing explanations tend to be static and based on the unitary actor assumption. This thesis attempts to make a step forward by relaxing the unitary actor assumption and by introducing dynamic elements into the explanatory framework. In order to do so, it takes a three-step approach, examining (1) the initial intensity of rebellion, (2) the onset of rebellion, and (3) the escalation. As for the methodology, this thesis adopts a mixed-methods approach: for each step, it conducts both large-N analyses and qualitative analyses of ex-Yugoslav countries. This thesis firstly examines the determinants of the initial intensity of rebellion. It presents two ideal-types, “decisions from above” made by ethnic leaders and “decisions from below” made by non-leaders, and will argue that the initial intensity is more likely to be higher if ethnic leaders organise the rebellion, because they can mobilise more resources. Secondly, this thesis examines the causes of the onset of rebellion. It will argue that structural conditions alone do not suffice to explain it, and will argue for the importance of dynamic grievances. In other words, even when structural conditions do not change, grievances will increase over time as the duration of peaceful protests gets longer, because people will be increasingly frustrated by the failure to achieve their goals by peaceful means. Finally, this thesis examines the dynamics of escalation. It will be argued that the low-intensity rebellion is likely to provoke state repression, and that the repressive measures taken by the state in turn are likely to cause the escalation of rebellion. The concluding chapter explores the theoretical and methodological implications for the study of ethnic conflict as well as the policy implications for conflict prevention. 3 Contents Introduction 10 Part I: Theoretical and Methodological Premises Chapter 1 A Three-Step Approach for the Analysis of Ethnic Rebellion Introduction 18 1. Definition of the Ethnic Rebellion 18 2. Four Problems: Critical Review of the Present Theoretical Literature 23 (1) Unitary Actor Assumption 24 (2) A Controversy - Do Grievances Matter? 27 (3) Lack of Dynamics 31 (4) Lack of Attention to the Escalation of Violence 33 3. Three-Step Approach for the Analysis of Ethnic Rebellion 36 (1) Initial Level of Rebellion 36 (2) Onset of the Initial Rebellion 38 (3) Escalation of the Rebellion 38 Conclusion 39 Chapter 2 Methodological Premises: Large-N analysis and the ex-Yugoslav Case Study Introduction 41 1. Methodology: Why Mixed Methods? 41 2. The MAR Dataset 44 3. The ex-Yugoslav Cases: Croatia, Bosnia, Serbia and Macedonia 49 4. Structure of the Thesis: Connection between Theory, Large-N Analyses and Case Studies 56 Conclusion 58 4 Part II: Initial Level o f Rebellion Chapter 3 Intra-ethnic Division, Decision-Makers and the Rebellion: Quantitative Analysis of the Initial Level of Rebellion Introduction 60 1. Two Ideal Types: Decisions From Above and Decisions From Below 61 2. Drawing Testable Hypotheses 68 3. Testing Hypotheses: Large-N Analyses 75 (1) Bi-variate analysis 76 (2) Multi-variate analysis 81 Conclusion 88 Chapter 4 Decisions from Above and from Below: Comparative Analysis of the Serb and Albanian Rebellions Introduction 89 1. Serbs in Croatia 91 (1) Involvement of Politicians and Police Forces in the Rebellion 91 (2) Involvement of the Regular Military Force in the Rebellion 95 (3) High Level of the Initial Intensity of Rebellion 97 2. Serbs in Bosnia 98 (1) Involvement of Politicians in the Rebellion 98 (2) Involvement of the Regular Military Force in the Rebellion 101 (3) High Level of the Initial Intensity of Rebellion 102 3. Albanians in Kosovo 104 (1) Path Chosen by the Albanian Elite in Kosovo: Peaceful Resistance 104 (2) Decision from Below: the KLA and Its Limited Resources 107 (3) Low Level of the Initial Intensity of Rebellion 109 4. Albanians in Macedonia 110 (1) Path Chosen by the Albanian Elite: Moderation and Participation 110 (2) Decision from Below: the NLA and Its Limited Resources 114 (3) Low Level of the Initial Intensity of Rebellion 115 Conclusion 116 5 Part III: Onset o f Initial Rebellion Chapter 5 Etiology of Ethnic Rebellion: Quantitative Analysis of the Effect of Dynamic Grievances Introduction 118 1. Theoretical Model: Dynamic Grievances and Structural Conditions 118 (1) Structural Conditions 121 (2) Dynamic Grievances: Importance of the Time Dimension 126 (3) The Effect of the Type of Decision-Makers 129 2. Large-N Analysis 131 (1) Research Design and Operationalisation 132 (2) Empirical Results 138 Conclusion 146 Chapter 6 Dynamic Grievances and Structural Conditions: Qualitative Analysis of the Ex-Yugoslav Countries Introduction 149 1. Albanians in Kosovo 150 (1) Structural Conditions 150 (2) Dynamic Grievances and the Availability of Arms 153 2. Albanians in Macedonia 159 (1) Structural Conditions 159 (2) Dynamic Grievances and the Availability of Arms 162 3. Serbs in Croatia 167 (1) Motivating Conditions 167 (2) Enabling Conditions 169 4. Serbs in Bosnia 172 (1) Motivating Conditions 173 (2) Enabling Conditions 175 Conclusion 177 Part IV: Dynamics of Rebel Escalations Chapter 7 State Repression as an Intervening Variable: Quantitative Analysis of Rebel Escalations Introduction 180 6 1. Theoretical Exploration 181 (1) Dynamics between the Ethnic Group and the State 181 (2) Structural Factors for State Repression 184 (3) Structural Factors for the Escalation of Rebellion 185 2. Large-N Analysis 187 (1) Research Design 187 (2) Empirical Results 195 Conclusion 203 Chapter 8 Repression or Conciliation? Comparative Case Study of Serbia and Macedonia Introduction 205 1. Case of Serbia: Military Repression and the Escalation 205 (1) State Reactions to the Initial Rebellion 205 (2) Effect of Repression: Expansion of the KLA 208 (3) Second-Phase Operations: Serbian Summer Offensive 212 (4) From the Holbrooke Agreement to Rambouillet 215 2. Case of Macedonia: Peace Process and Prevention of Escalation 217 (1) Trajkovski’s Political Process for the Solution of Conflict 218 (2) Negotiations at Ohrid 220 (3) Importance of Trajkovski’s Political Process 223 Conclusion 225 Chapter 9 Explaining State Reactions: Path Dependency and External Pressures in Serbia and Macedonia Introduction 226 1. Explanatory Factors: Path Dependency and External Pressures 227 2. Case of Serbia 229 (1) Path-dependency and the Context of Domestic Politics 229 (2) Ambiguous Signals from External Actors 235 3. Case of Macedonia 239 (1) Path Dependency: Election with the Support of Albanians 239 (2) External Pressures from the International Community 243 Conclusion 248 7 Conclusion 1. Key Findings of the Thesis 251 2. Implications of the Case Selection 253 3. Theoretical and Methodological Implications 257 (1) Unitary Actor Assumption 257 (2) Conceptualisation and Operationalisation of Civil War 258 4. Policy Implications 260 5. Tasks for Future Research: Additional Implications and Theoretical Extensions 262 Reference 265 Appendices 283 Tables and Figures Tables 1.1 Initial Level, Onset and Escalation 37 3.1 Type of Decision-makers and the Initial Level of Rebellion 66 3.2 Status of the Ethnic Group within the State and the Initial Level of Rebellion 67 3.3 Type of decision-makers, the Status of the Ethnic Group and the Initial level of the rebellion 68 3.4 Initial Level of Rebellion and Involvement of Politicians, 177 ethnic groups from 1945 to 2000 78 3.5 Initial Level of Rebellion and Involvement of Police or Army, 177 ethnic groups from 1945 to 2000 80 3.6 Ordered Logit Analysis of the Initial Level of Rebellion, 177 ethnic groups 85 5.1 Random-Effects Logit Analysis of Determinants of the Onset of Rebellion, 1945-2000 (Dependent Variable: REBONSET) 139 5.2 Random-Effects Logit Analysis of Determinants of the Onset of Rebellion, 1945-2000 (Dependent Variable: REBONSET2) 140 5.3 Random-Effects Logit Analysis of Determinants of the Onset of Rebellion, 1945-2000 (Dependent Variable: